← All documentation
Guide 01 — Node Setup

Running a Node

Dolphin Network node quick-start — spin up an inference worker and start earning POD.

  1. 1

    Visit the Dolphin Network app

    Open the Dolphin Network app at https://v2.dphn.ai.

  2. 2

    Authenticate with your wallet

    Authenticate with wallet on the Dolphin app
  3. 3

    Create a key in the "API Keys" section

    Create an API key
  4. 4

    Press "worker link"

    Press the worker link button
  5. 5

    Copy the script to run the worker

    Copy the worker bootstrap script
  6. 6

    Run this command on your GPU server

    Paste and run the worker command on your GPU server
    60 GB vRAM required to run in FP8 with full context.

    Suitable GPUs:

    • 1x RTX 6000 PRO
    • 1x H100 / A100 80 GB
    • 4x 3090 / 4090
    • 2x 5090
    • 2x A6000
    • 1x H100 / B200 / B300

    We recommend Targon for GPU rentals.

    If using Targon, forward port 9344 in the config before starting the instance.

    Forward port 9344 in the Targon config

    This is used to remotely view your worker through the WebUI.

  7. 7

    View your worker status remotely

    Once the worker has loaded you will see this screen.

    Worker WebUI status screen

    You may use this IP to view the worker's status remotely from any browser.

    For Targon users

    Targon service ports section

    Take the IP from the service ports section on Targon's website and add /?token= followed by your token to the end of the URL.

    For example:

    https://wrk-tbwbogod5h51-9344.caas.targon.com/?token=98C6537ED7604A42CCA5B5BBA37277D5